os9@nyit.UUCP (OS9 News) (01/15/86)
Mod.os.os9 Digest Number 1, 15-January-1986
Contents:
1. Welcome to mod.os.os9 (nyit!os9).
2. QT+ 68000 review by Bob Larson (oberon!blarson).
3. Some ideas from Jim Omura (lsuc!jimomura).
--------------------------------------------------------------------
From: nyit!os9 Wed Jan 15 13:00:00 1986
Subject: Welcome to mod.os.os9!
Welcome to the mod.os.os9 newsgroup!
This newsgroup is available for discussion on OS-9 at any
technical level. The moderators will route questions to
knowledgeable OS-9 programmers, who will post the questions
with answers. Those of you who would like to answer questions
should send mail to the moderators, listing the areas in which
you are proficient.
An OS-9 sources archive is being created at NYIT. A table of
contents will be posted regularly, and sources will be distributed
by mail on request.
To post to the newsgroup, MAIL your message to nyit!os9 (see
routing info below). Other correspondence about the newsgroup
should be mailed to nyit!os9-request.
The moderators are Alex Arthur (nyit!aca), and Bruce Perens
(nyit!bp) of the NYIT Computer Graphics Laboratory. We'd like
to thank Jim Jones (uokvax!emjej), Mark Horton (cbosgd!mark),
Gene Spafford (gatech!spaf), and Dave Mason (mason@toronto.CSnet)
for their help in creating the newsgroup.
--------------------------------------------------------------------
OS-9 is a trademark of Microware.
516-686-7644 (Bruce Perens or Alex Arthur)
{allegra,seismo,decvax,vax135,ihnp4,mcvax}!philabs!nyit!os9
nyit!os9%suny-sb.CSnet@CSnet-Relay.ARPA
--------------------------------------------------------------------
From: Bob Larson <oberon!blarson> Tue Jan 14 19:55:47 1986
Subject: QT+ review
Keywords: QT+, 68k, hardware review, FHL
Overview:
The QT+ is a 10 Mhz 68000 system that runs os9/68000 (known as osk).
My system has 1/2 Megabyte of memory (additional 1/2 megabyte expantion
available) and two DSQD floppy drives. I plan on adding hard disk, but
have not yet done so. The Frank Hogg Labs QT+ is based on the Hazelwood
Uniquad II single board compuer, I posted a comparison of it with a couple
of other os9/68000 single board computers (based on literature and seeing
them at the os9 conference) to info-68k and net.micro.68k a couple of
months ago. (If needed, I could mail you a copy.)
My system arrived from FHL about two weeks after I ordered it. (The delay
may have been due to paying via personal check.) Since I ordered it at the
show, it included a clear plastic top which can be used in place of the
normal case top to show off the insided without getting them dusty.
Unfortunatly it arrived with a defective boot disk, due not to a media defect
but an error creating the master. In responce to my complaint, Frank Hogg
mailed out a bootable disk which I received in 3 working days. (I don't
think they will make this mistake again.)
Documentation:
The documentation covering os9/68000 is quite good. It does have a number
of typo's and minor mistakes, but nothing that I found is intentionaly
misleading or obviously missing.
The documentation of stylograph assumes you have never seen a termial
before. I havn't yet had the patiance to wade through it and see if there
is a reasonable word processor burried under there or not. (Three pages on
what the control key is is a little more than I was interested in at the time.)
The dynacalc manual is small and looks like a reference guide rather than
an introduction. I havn't yet tried dynacalc.
The lack of good hardware documentation is a problem. FHL includes 16 pages
of QT+ documentation, (not counting covers) some of which refers to the 68008
processor of the QT. Things that I think should be included are a schematic,
wiring of the serial ports and a better description of both the serial and
parrelell ports, better desciptions of the jumpers and what they do, details
on the SCSI port (sometimes refered to as SASI, confusing to the uninitated),
how the timers are programed, desription of floppy disk i/o, and I'm just
scratching the top.
Most of the documentation fits in small notebooks, one is included but that
is not sufficent for the documentation suppled. I got another notebook with
the C compiler that had sufficent room for the overflow. The dynacalc
documentation comes in a separte little binder. The FHL QT+ documenation
comes on loose 8.5*11 paper, it isn't even stapled or punched.
Hardware:
The QT+ comes in a nice small case (probably designed for an external hard
dive for something else). The rear pannel has connectors for power, two
parrellel printers (db25p), four serial devices (db25s), and external floppies.
Internally there are connections for SCSI and extra I/O which are not brought
to the rear pannel. I'll proabably add a connector for an external SCSI
controler.
The fan is noiser than I would like. I may replace it with a quieter one
with the same dimentions. The system does run cool. It appears that only
the fan, fuse and a jumper on the power supply would need to be changed for
operation at 220 volts.
The serial lines use jumpers to select their speed. This is fine for
hard wired terminals and printers but isn't good for a multi-speed modem
or port selector. The uniquad and 68B50 documentation talks about the
ports as if they were wired for DTE while they are wired as DCE. 2, 3, 6, 7,
and 20 are the live pins, which I found out by examining the wireing inside
the machine.
The documentation shows how to wire a cable for the parellel printer but
does not state if it is identical to the IBM PC cables that use the same
connectors and are readily available. I am currently using a serial printer
so I havn't tried the parrelle ports.
The floppy interface has the annoying habit of dissabling interupts durring
floppy transfers. This slows file transfer programs like Kermit (which I have
working) and could get annoying on a multi-user machine. Coco disks can be
read but not formatted. This means that you will need a 40-track drive or a
coco with a 80-track drive to write coco disks reliably. (A disk formated
by a 40 track drive cannot be reliably writen to by a 96-track drive.)
The boot roms (2058 of 4096 bytes used) contain an undocumented memory dump
routine. Answer @ to the boot device prompt to get to it, and specify two
six digit hex addresses (from and to). Use reset to exit. Part of osk
could be moved to eprom to reduce ram usage if desired.
Software:
The osk operating system is a bit more complete than coco os9 is. A
relocating macro assembler, a make utility, and a tolerable screen editor
are included. (The screen editor could be used with most, but not all,
terminals that I know of.)
The C compiler, separatly purchased (list $400), is slower than I had hoped.
It does fix a couple of problems in the 6809 version, but adds at least one
non-standard annoyance (remote is a reserved word) and isn't bug free. (I've
had problems with the -d compiler option.) Using ram disk for the temporary
files does help some with the speed.
I havn't realy tried Basic09, Stylograph, or Dynacalc yet. None of the
promised additional software from FHL has arrived yet. I did get kermit
working, the new version with compiler options for both os9/6809 and osk
will be standard when I get confirmation I didn't break the 6809 part.
--
Bob Larson
Arpa: Blarson@Usc-Ecl.Arpa
Uucp: ihnp4!sdcrdcf!oberon!blarson
--------------------------------------------------------------------
From: lsuc!jimomura Tue Jan 14 14:23:58 1986
Subject: A few ideas and some self-congratulation from Jim Omura!
Hi Guys:
First of all, congratulations to us all! By backing OS-9 or showing
interest in it, we're showing that we are interested in good software
technology and not willing to blindly follow the masses. At any given
time it's hard to say what the "forefront" of our technology is. It
advances in so many ways, but, the argument that something is best just
because it's the most popular has always been specious. Now we can
at least hope that others will learn from us as we learn from each other.
Now for some nitty gritty. There are a couple of ideas that I've
been thinking about just recently. They may have been dealt with before
by the Users' Group (of which I'm not currently a member) but I was wondering
if there was any consensus.
1. Termcaps -- I was looking at the original "microEMACS" source-code
and I can see that it wouldn't be hard to use a "termcaps" file
system. Is there any established practice as to where "termcaps"
files go in OS-9? I'm personally leaning towards sticking it in
the "SYS" directory.
2. Does anyone have the sourcecode for a public domain formatter to
go along with "microEMACS"? I think the Users' Group has one
called "print" but I'm not sure. I'm hoping to find a public
domain version of 'nroff', but it doesn't have to be.
3. While discussing Modula-2 on BIX, I came up with the idea of
creating a common-block handling system for OS-9. The modules
would be a 'cbman' (common block manager) and 'cb' device
descriptor. This would be an optional set of modules used
for languages which could support common blocks for data
structure sharing between free-standing modules. Has anyone
else considered this?
4. LANS -- Has anyone done any work regarding LANS yet?
I currently have the source code for Bob Santy's OS-9 version of
'microEMACS'. There's only one problem. He wrote it on an 80 track
drive. I only have 40 track drives. As such, I can't read the format.
As soon as I can find someone to help me with this, I'll be posting the
source on BIX and, subject to his permission (which I'm fairly certain
he'll give) I'll post to the net at about the same time.
[moderator's note - Al Arthur can help if Jim can't find anyone
closer. -bp]
Cheers! -- Jim O.
END OF OS-9 DIGEST
--
OS-9 is a trademark of Microware.
516-686-7644 (Bruce Perens or Alex Arthur)
{allegra,seismo,decvax,vax135,ihnp4,mcvax}!philabs!nyit!os9
nyit!os9%suny-sb.CSnet@CSnet-Relay.ARPA